Moral hazard, asset specificity, implicit bonding, and compensation: the case of franchising.
Wimmer, Bradley S. ; Garen, John E.
I. INTRODUCTION
There is a large and growing interest in the structuring of
compensation to deal with the moral hazard problem. Much of this
literature treats various forms of income sharing, such as piece rates
and sharecropping, as a means for a principal to provide incentives to
an agent. This literature has been extended to the case of two-sided
moral hazard where both parties to an exchange may shirk. Another
mechanism to induce effort discussed in the literature is bonding, where
workers post a bond that is forgone if effort is inadequate. Klein
[1980] and Klein and Leffler [1981] treat the possibility of specific
assets as constituting an implicit bond, where poor performance by a
firm is penalized by bankruptcy and sale of assets at a loss.
This paper studies a situation where these issues coalesce: the case
of compensation arrangements in franchising. The typical franchise
contract calls for the sharing of franchise income between franchisee
and franchisor. Additionally, franchisees may be terminated for poor
performance. If franchise assets have a degree of specificity, the
ensuing sale of assets at a loss provides a penalty for poor
performance.
We examine the determinants of franchise royalty rates and fees in
this setting. While there is a literature on two-sided moral hazard and
some empirical work on franchising based on this idea, none incorporates
the potential influence of asset specificity. We explicitly consider the
role of implicit bonding and asset specificity in conjunction with moral
hazard on the part of both trading parties. We go beyond previous work
in franchising by estimating the effects of asset specificity and other
influences on the royalty rate. Further, part of our empirical work
utilizes improved micro data on actual franchise contracts.
Our approach and its implications are discussed in section II.
Consistent with previous work, we expect the royalty rate to decrease
(increase) as the importance of the franchisee's
(franchisor's) input increases. However, more specific capital,
with the threat of termination, acts as an implicit bond and substitutes
for a reduced royalty rate in inducing franchisee effort. Thus, we
expect royalty rates to increase with asset specificity. Also, unlike
other approaches, we do not necessarily expect an inverse relationship between the royalty rate and the franchise fee. For example, the greater
the importance of franchisor effort the higher the royalty rate, but the
franchise fee may also increase if franchise revenue rises by enough.
Our hypotheses are tested using two different data sets. An original,
firm-level data set is taken from Federal Trade Commission
(FTC)-required Uniform Franchise Offering Circulars (UFOC).(1) In
addition, data are obtained from Bond's 1989 The Source Book of
Franchise Opportunities.
The data are discussed and the results presented in section III. We
find substantial support for our approach. For example, factors that
increase the importance of the franchisee's effort, such as the
number of employees supervised, lower the royalty rate and increase the
franchise fee. Also, a reduction in the specificity of the
franchisee's investment due to leasing lowers the royalty rate and
raises the franchise fee.(2) Section IV summarizes our findings and
concludes the paper.
II. THE APPROACH AND THE HYPOTHESES
The typical franchising contract consists of a two-part payment by
the franchisee to the franchisor. The franchisee pays an up-front,
nonrefundable franchisee fee and a continuing royalty payment, usually a
percentage of sales. In return the franchisee receives the right to use
the franchisor's brand-name capital and a method of operation. In
addition, the franchisor provides on-going support to the franchisee.
However, the franchisor maintains the right to unilaterally terminate
the contract if the franchisor determines that the franchisee is not
delivering a product consistent with the quality guaranteed by the brand
name. Our goal is to explain variations in the royalty rate and
franchise fee in this context.
Denote the franchise fee as F and the royalty as R. We follow the
literature in recognizing (e.g., Rubin [1978]) that, because franchisors
and franchisees provide inputs to the franchise, R will be set to
account for the effect on both franchisee and franchisor incentives.(3)
In particular, a greater (lower) R gives the franchisor (franchisee)
more effort incentives because (s)he retains a larger share of the
returns to effort.(4) Lafontaine [1992] and Sen [1993] find empirical
support for the idea that moral hazard considerations contribute to
explaining variations in royalty rates.(5)
In our work, we also recognize that implicit bonding plays a role in
influencing the franchise payment structure.(6) Mathewson and Winter
[1985] treat explicit bonding and its difficulties in franchising. In
another context, Klein and Leffler [1981] argue that specific assets may
constitute an implicit bond. If the owner of the specific assets goes
out of business due to poor performance, the specific assets can be sold
only at a loss, thus providing incentives to perform. Klein [1980]
recognizes that this incentive mechanism may operate in franchising.
Here, we explicitly incorporate how asset specificity creates an
implicit bond and affects the use of the royalty rate to deal with
two-sided moral hazard.(7) The simultaneous consideration of these
issues makes our approach unique in the literature.
The potential of termination for poor performance generates the
incentive of the implicit bond. If termination occurs, franchisees sell
the assets for their salvage value. Assume the assets are purchased for
I and have salvage value [Alpha]I, where [Alpha] [less than or equal to]
1. Thus, 1 - [Alpha] is the degree of asset specificity. The amount (1 -
[Alpha])I is forgone if termination occurs and so is equivalent to a
bond. Higher franchisee effort can improve performance and reduce the
chances of termination. Thus, the more specific are franchise assets,
the higher is the implicit bond and the greater is franchisee effort.
Note that under certain conditions asset specificity may have the
opposite effect on franchisor effort incentives. Terminated franchise
assets can be purchased for [Alpha]I but have value of at least I in the
franchise system. Thus, there may be a temptation for franchisors to
reduce their support to franchises to induce poor performance and
termination. This way the franchisor might purchase the assets for
[Alpha]I and redeploy them elsewhere within the franchise system for
benefit I. How attractive this temptation is depends on the desire of
the franchisor to maintain a good reputation and the ease of inducing
termination. The point remains that, all else constant, greater asset
specificity may reduce franchisor effort incentives.
In designing the franchise contract, the franchisor establishes a
royalty rate ex-ante taking account of the above discussed ex-post
incentives.(8) The wealth-maximizing choice of the royalty rate trades
off franchisor and franchisee effort, i.e., a higher R raises the former
and reduces the latter. A straightforward prediction of this approach,
also found elsewhere in the literature, is that the greater is the
franchisor's (franchisee's) marginal product of effort, the
higher (lower) is the royalty rate. Asset specificity also plays a role.
Greater asset specificity raises franchisee effort, making it optimal to
use the royalty rate to generate more franchisor effort.(9) Thus, we
predict asset specificity will raise royalty rates. This is a unique
prediction in the literature.
Franchisor reputation has an ambiguous effect on the royalty rate in
our approach. A better reputation may raise the marginal product of both
parties and presumably also reduces the franchisor's incentive to
wrongfully terminate franchises.(10)
Consider now effects on the franchise fee. In our approach, the
franchise fee is a rent-transfer device, bringing the franchisee down to
his or her reservation level of utility and having no allocative
effects. An exogenous variable affects the franchise fee in two basic
ways: (1) if it causes total franchise income to rise (fall), the
franchise fee rises (falls); (2) if it causes the franchisee's
share of franchise income, 1 - R, to rise (fall), the franchise fee
rises (falls). These influences may reinforce or offset one another.
A increase in 1 - [Alpha], the specificity of the franchise assets,
causes the royalty rate to rise and reduces the value of the franchise
in the event of termination. Both tend to reduce the franchise fee. An
increase in I, [Alpha] constant, increases the amount of specific
investment thus tending to reduce the franchise fee. However, a greater
investment raises total franchise income, increasing the fee. Its
overall effect is ambiguous.
An increase in the marginal product of franchisee effort reduces the
royalty rate and raises F. If the higher marginal product also is
associated with a higher total product, franchise income is higher also
causing F to rise. An increase in the marginal product of the
franchisor's effort raises the royalty rate and so lowers the fee.
However, if associated with a greater total product, thereby raising
franchise income, this raises the fee. No clear prediction can be made.
Interestingly, we do not find that the royalty rate and the franchise
fee necessarily move in opposite directions, as some authors claim
(Blair and Kaserman [1982] and Lafontaine [1992]). While factors that
raise the royalty rate tend to lower the fee, there are other effects.
Franchisor reputation likely increases the revenue of the franchise
directly, suggesting a higher fee. However, reputation has an unclear
effect on the royalty rate, creating some ambiguity.
III. THE DATA AND EMPIRICAL TESTS
Data Description
To test the hypotheses discussed above, two data sets were compiled.
Data from The Source Book of Franchise Opportunities, 1989 Edition by
Robert E. Bond (the Bond data set) and an original data set gathered
from FTC-required UFOC are used to construct variables that proxy the
above described effects.
The Source Book is a publication that aids the prospective franchisee
in the selection of a franchisor and contains a large number of
observations covering a wide range of franchising activity.(11) Our
sample is limited to franchisors based in the United States that rely
primarily on royalty payments and franchise fees for revenue. Based on
this criterion, and after deleting observations with missing data on key
variables (see below), 528 observations remain and are included in the
analysis.
A UFOC provides potential franchisees with detailed information on
the franchise contract and certain aspects of the franchisor. The UFOC
data was obtained through visits to the state of Indiana's
Securities Department and a general mailing to franchisors.(12) In order
to utilize information from both sources of data, UFOC franchises were
matched to the Bond data set. The final UFOC data set includes 196
franchisors; 131 franchisors that were on file with the Indiana
Securities Division during the period February 1990 to August 1990 and
65 franchisors from the general mailing.
The royalty payments are reported as the fraction of franchisee sales
paid to the franchisor. Royalty payments typically include both a
royalty rate and an advertising contribution, the latter to be used
solely for advertising. The empirical analysis includes regressions that
use the royalty rate by itself and the sum of the royalty rate and the
advertising contribution, denoted the share, as dependent
variables.(13,14) The franchise fee is the initial, nonrefundable
payment made by the franchisee to the franchisor. Descriptions, means,
and standard deviations of all variables, along with predictions for the
independent variables, are found in Table I.
A unique feature of our approach is the expectation that an increase
in a franchisee's asset-specific investment increases the royalty
rate. The franchisee's initial investment, net of the franchise
fee, measures the size of the franchisee's initial capital
requirements and is available from both sources of data.(15) Franchisee
investment, however, does not capture the investment's specificity.
We include information about the "type" of structure that the
franchisee is required to obtain in an attempt to proxy asset
specificity. In the UFOC data, dummy variables indicating if franchisees
are required to build or purchase a structure, are allowed to lease, or
may run the franchise from their homes are included in the empirical
analysis.(16)
The type of structure depends largely on the franchise and often
includes features unique to the franchising operation. When the
franchisor requires the franchisee to acquire a unique structure, any
necessary remodeling of an existing structure requires a large degree of
franchisee-specific modifications. We expect franchisees to build or
purchase the structure when the investment is highly specific because
the nonrenewal of a lease places a large capital loss on the franchisor
(Smith and Wakeman [1985]). When leasing premises, franchisees are not
under the burden to sell the capital should the franchising contract be
terminated, and this implies a lower degree of asset specificity. When
franchisees are allowed [TABULAR DATA FOR TABLE I OMITTED] to run the
franchise out of their homes, the investment may be highly specific to
the franchise but specificity is expected to be reduced by the lack of
site specificity. Thus, a building requirement is expected to have a
positive (negative) effect on the royalty rate (franchise fee) while
working out of the home is expected to have the opposite effect,
assuming that leasing is the omitted case. Additionally, a dummy
variable identifying franchisees who lease capital equipment is included
and should decrease (increase) the royalty rate (franchise fee).
The effects of the level of investment and its degree of specificity
on royalties are examined by entering investment and its interactions
with the structure-type dummies. Investment is expected to have a
positive effect on the royalty rate, with the effect being larger the
greater its specificity. Thus, the investment variable in the lease case
is predicted to have a positive effect on the royalty rate. We expect
the interaction of the building dummy with investment to be positive as
specificity is greater in this case. The lack of site-specificity in the
home case implies that there is no clear-cut prediction for the
interaction of the home dummy with investment. In the franchise fee
regressions, recall that there is no unambiguous prediction concerning
the effect of investment on the franchise fee.
The Bond data set has less detail regarding the nature of the
franchisee's investment. While the level of investment is
available, we are able only to discern if the franchisee does not lease
capital. A nonlease dummy variable is created to indicate this. It does
not specify whether the franchisee must construct an outlet or may
operate out of the home. The absence of leasing indicates an increase in
asset specificity and so the nonlease dummy and its interaction with
investment should raise the royalty rate and the lack of leasing itself
should lower the franchise fee.
An important hypothesis of our approach and elsewhere in the
literature is the effect of the franchisee's and franchisor's
marginal products of effort on the terms of the franchise contract. The
franchisee's marginal product of effort is proxied by the average
number of employees and the productivity of each employee. Employee
productivity is likely to be inversely related to the labor-capital
ratio, proxied here by employees over investment. Accordingly, the
royalty rate is predicted to be negatively related to the number of
employees and positively related to the labor-capital ratio. The reverse
holds for the franchisee fee.
Also, an increase in the duration of franchisee training suggests an
increase in the marginal product of franchisee effort. However, to the
extent that training is franchise specific, training also may proxy for
the amount of the franchisee's investment in specific human
capital. It follows that no unambiguous prediction can be made regarding
the effect of training.
It is expected that an increase in the advertising contribution
implies an increase in the importance of national brand-name capital and
the franchisor's marginal product of effort. This effect is
reinforced as the number of franchised outlets increase. We expect the
size of the advertising contribution and number of outlets to be
positively related with the royalty rate, but to have an ambiguous
effect on the franchise fee.
We also account for other potential influences on the royalty rate
and franchise fee. To control for horizontal free riding and geographic
dispersion, we include a dummy variable indicating if the franchisee
operates in a sector characterized by nonrepeat-purchase customers and
the number of states franchised, respectively. The nonrepeat dummy
follows Caves and Murphy's [1976] taxonomy.(17) The number of
states franchised is expected to be positively related to the cost of
franchisor inputs (see Rubin [1978], Brickley and Dark [1987], and
Lafontaine [1992]).
We also include as control variables proxies for monitoring, risk,
and reputation (see Lafontaine [1992]). Klein and Saft [1985] argue
monitoring by franchisors is easier if franchisees are required to
purchase inputs directly from the franchisor. This is proxied by dummy
variables indicating the presence of a franchisee purchase requirement.
As in Lafontaine [1992], we proxy the franchisor's reputation with
the number of years the franchisor operated before franchising the
system and the number of years the franchising system has been in
operation. To measure risk, the coefficient of variation of sales was
created for each industry using industry average sales data from
Franchising in the Economy [U.S. Department of Commerce 1988, 1990] for
the years 1976 through 1988.(18,19)
The Results. The empirical analysis consists of tobit estimates of
the royalty rate and franchise fee on the variables described above and
in Table I. The findings are shown in Table II for the UFOC data and
Table III for the Bond data.
Consider first the findings regarding the proxies for asset
specificity with the UFOC data in Table II. An increase in investment,
holding the degree of asset specificity constant, is expected to
increase the royalty rate. The investment variable, which corresponds to
the lease case, is positive although insignificant. The effect of the
building and home dummies is revealed by their coefficients and their
interactions with investment. For the building case, the dummy is
positive, as expected, and its interaction is negative, which is
contrary to our expectations. Both are significant. Using the
coefficients in column 1, the net effect is that the royalty rate is
higher when the franchisee must build as long as investment is less than
$1.77 million. This holds for 94% of our sample. For the home case, the
dummy is negative and its interaction is positive, with both attaining
significance. However, computing its net effect on the royalty rate with
the coefficients of either column, we find it is negative for two-thirds
of the sample. Additionally, the lease capital equipment dummy has a
negative and significant effect. Overall, these results are in general
agreement with our prior that an increase in the size of the
franchisee's specific investment leads to a higher royalty rate.
We also expect that asset specificity lowers the franchise fee, but
that total investment has an ambiguous effect. In the UFOC data, we find
the effect of the building dummy in conjunction with its interaction is
negative in both columns 3 and 4, as expected. Only the interaction is
significant. Also, the lease capital equipment dummy has a positive
effect, as predicted, but it is not significant. Working out of the home
tends to reduce the fee, contrary to expectations, although its
significance is not overly strong.
The Bond data set has less detailed proxies regarding asset
specificity: only the nonlease dummy and the investment variables.
Referring to Table III, the effect of the nonlease dummy variable, when
combined with its interaction with investment, is positive and
significant in the royalty rate equation, as expected. In the share
equation, the interaction term is positive and significant, as expected,
but the dummy is negative although not significant. The positive and
significant coefficients for the interactions of nonlease and investment
in both the royalty rate and share equations indicate that the royalty
rate increases with the level of investment for franchisees [TABULAR
DATA FOR TABLE II OMITTED] that do not lease. The investment variable
itself is negative but insignificant. The prediction that investment
matters less when the franchisee is allowed to lease capital holds.
Regarding the findings for the fee using the Bond data, the nonlease
dummy is negative and significant and its interaction with investment is
positive and significant. The former [TABULAR DATA FOR TABLE III
OMITTED] is supportive of our approach while the latter is not.
The results regarding investment and its degree of asset specificity
are broadly consistent with our approach. As in Lafontaine [1992], we
find that both the royalty rate and franchise fee are affected by the
size of the franchisee's capital requirements. However, the size
and direction of this relationship depend on the type of investment made
by the franchisee. While it is difficult to proxy the degree of asset
specificity, our findings are consistent with the hypothesis that
leasing reduces the degree of asset specificity and the size of the
franchisee's implicit bond, inducing a lower royalty rate and
higher fee.
As in Lafontaine [1992] and Sen [1993], our empirical analysis
supports the hypothesis that the marginal product of both the
franchisee's and franchisor's efforts are important
determinants of the royalty rate. The coefficient on the number of
employees is negative in the royalty rate and share equations for both
data sets. It is significant, however, only for the UFOC data. The
coefficients for the employees-investment ratio are positive and
significant in all regressions reported. The advertising contribution
coefficient is positive for both data sets, reaching statistical
significance in only the UFOC data. The number of outlets franchised is
positive, as expected, and significant in all the royalty rate and share
regressions.
As predicted, an increase in the number of employees raises the
franchise fee and a rise in the employee-investment ratio lowers it.
Statistical significance is reached only with the Bond data set. While
we have no predictions regarding the sign of the advertising or number
of outlets variables in the franchise fee equation, the former is
positive, although insignificant, in both data sets and the latter is
negative and significant.
The training variable is negative in both data sets for the royalty
rate regressions. A similar result is found for the share regressions,
although statistical significance is low. In the fee regressions,
training is positive and significant in all specifications reported.
These results are consistent with a longer training period increasing
the franchisee's marginal product of effort.
The remaining variables provide mixed results. The number of states
franchised, proxy-ing geographic dispersion, is insignificant in all
specifications reported. In the fee regressions, this variable is
positive and significant for all specifications. The nonrepeat customer
variable, proxying for horizontal free riding, has negative coefficients
in the royalty rate and share regressions, but is significant for only
the Bond data. The influence on the franchise fee tends to be negative
and significant.
The results concerning the monitoring, reputation, and risk proxies
are generally consistent with those of Lafontaine [1992] and Sen [1993].
Input purchase requirements tend to reduce the royalty rate and have
little effect on the fee. An increase in the franchisor's
experience tends to raise the royalty rate, and an increase in the years
the franchise system has operated reduces the fee in some
specifications. The coefficient of variation of sales tends to raise
both the royalty rate and the fee, although significance is mixed.
IV. SUMMARY
The case of franchising is one where many of the elements of moral
hazard models merge. Issues of two-sided moral hazard, bonding, and
asset specificity all play a role. The typical franchise contract calls
for the sharing of franchise income between franchisee and franchisor,
thus the issue of the incentive effects of residual income claimancy
arises. Additionally, franchisees may be terminated for poor
performance. If franchise assets have a degree of specificity, the
ensuing sale of assets at a loss provides a penalty for poor performance
and acts as an implicit bond. Therefore, bonding plays a role in the
franchise contract.
Our work examines the determinants of franchise royalty rates and
fees in this setting. While there is a literature on two-sided moral
hazard and some empirical work on franchising based on this idea, none
incorporates the potential influence of asset specificity. We extend
existing literature by considering how asset specificity and bonding
interact with other determinants of incentive pay. Our empirical work
tests the hypotheses of this approach with both previously used and new
franchising data.
Many of our findings are consistent with the earlier work of
Lafontaine [1992] and Sen [1993]. For example, we find that the
franchisor's residual income claim - the royalty rate - rises
(falls) as the importance of the franshisor's marginal product
rises (falls). However, our work goes beyond this to show that asset
specificity and implicit bonding play a significant role. More specific
capital, with the threat of termination, acts as an implicit bond and
substitutes for a reduced royalty rate in inducing franchisee effort.
Thus, higher royalty rates emerge with greater asset specificity. The
general implication is that if one party to an exchange posts a larger
bond, this improves their incentives and allows enhancement of the other
party's incentives through a larger residual income claim. Our
empirical findings lead to the conclusion that incentive and bonding
issues are important in the determination of the compensation of
franchisees and franchisors.
The views expressed here are those of the authors and do not
necessarily reflect those of the Federal Communications Commission. We
are especially grateful to Frank Scott for many helpful discussions and
much valuable input. Also, the anonymous referees were most helpful with
numerous, detailed comments. We also wish to thank David Baucus, Mark
Berger, Glenn Blomquist, Richard Jensen and the participants in the
Applied Microeconomics Workshop at the University of Kentucky for
helpful comments. We thank the Indiana Securities Division and Frank
McCormick for help in data collection.
1. The Franchise Offering Circular discloses the franchise contract
to franchisees, outlining the obligations of all parties.
2. Other features of franchising, such as the use of company-owned
stores, are not examined here.
3. Other models have been put forth to explain sharing contracts,
including downstream rent extraction (Blair and Kaserman [1982] and
Schmidt [1994]), risk sharing (Cheung [1969] and Stiglitz [1974]),
measurement costs (Barzel [1982] and Leffler and Rucker [1991]), and
signaling (Gallini and Lutz [1992]). Lafontaine [1992] provides an
overview of this literature.
4. Two-sided moral hazard has been formally modeled in a franchising
context by Lal [1990] and Bhattacharyya and Lafontaine [1995]. Examples
of other applications include Eswaran and Kotwal [1985], Carmichael
[1983], Cooper and Ross [1985], and Mann and Wissink [1988].
5. Also, much empirical work on franchising finds that agency costs are important in determining the fraction of company-owned outlets. As
examples, see Brickley and Dark [1987], Lafontaine [1992], Norton
[1988], and Scott [1995]. Also, Martin [1988] examines risk in
determining company ownership of stores.
6. The issue of bonding was first formally treated by Becker and
Stigler [1974] and extended by Lazear [1979]. The infeasibility of
bonding is important in the efficient wage literature. For reviews and
assessments of this literature, see Carmichael [1990] and Lang and Kahn
[1990].
7. Minkler and Park [1994] examine the effect of specificity on the
vertical integration of franchises.
8. This assumes that there are noncontractable aspects to both
parties' efforts.
9. As noted previously, greater specificity also may reduce
franchisor effort, so the royalty rate is raised to offset this.
10. Some have argued that, because a better reputation raises sales,
the franchisor can charge a higher royalty rate (Blair and Kaserman
[1982]). Our analysis implies that this need not be true as changing the
royalty rate distorts incentives. In the context of a life-cycle model,
Mathewson and Winter [1985] show that established (and presumably more
reputable) franchises can have a higher royalty rate and a higher fee.
11. Sen [1993] and Scott [1995] also use data from The Source Book.
12. Indiana Disclosure Law IC 23-2-2.5 requires franchisors who wish
to operate in Indiana to register and file a UFOC with the state.
Franchises who meet equity requirements are exempt. To supplement the
Indiana data, a general mailing was sent to franchisors requesting a
UFOC. A response rate of 12.88% was obtained from the general mailing.
For more details regarding the Uniform Franchise Offering Circular, see
Wimmer [1992].
13. The UFOCs report, for some franchises, that the royalty rate may
vary over time or with franchisee sales. Dummy variables were created
for each of these cases and were included in specifications of the model
not reported. Their coefficients proved to be insignificant while not
affecting the other results. For the Bond data, when a range of royalty
rates is reported, the average is used.
14. The share variable is used by Lafontaine [1992] and Sen [1993].
The model is tested using the advertising contribution as a part of the
royalty payment and as an independent variable that proxies the
franchisor's marginal product of effort.
15. Lafontaine [1992] uses initial capital requirements to proxy
franchisor credit constraints.
16. The home case is a situation where franchisees may run the
franchise out of their home or may obtain separate premises to run the
business.
17. Caves and Murphy [1976] posit that repeat purchases are
unimportant in the restaurant, automobile rental and the hotel, motel,
and campground sectors.
18. For use of a similar variable, see Martin [1988], Norton [1988],
and Lafontaine [1992].
19. In the UFOC franchise fee regressions, the length of the
franchise contract, dummy variables indicating the presence of a
lifetime franchise contract and a franchise fee that varies according to market size are included. For the Bond data set, only a contract length
variable is available. These variables influence the present value of
franchisee's expected income but not effort incentives.
REFERENCES
Barzel, Yoram. "Measurement Cost and the Organization of
Markets." Journal of Law and Economics, April 1982, 27-48.
Becker, G. S., and G. J. Stigler. "Law Enforcement, Malfeasance,
and the Compensation of Enforcers." Journal of Legal Studies,
January 1974, 1-17.
Bhattacharyya, S., and F. Lafontaine. "Double-Sided Moral Hazard
and the Nature of Share Contracts." Rand Journal of Economics,
Winter 1995, 761-81.
Blair, Roger D., and David L. Kaserman. "Optimal
Franchising." Southern Economic Journal, October 1982, 494-505.
Bond, Robert E. The Source Book of Franchise Opportunities, 1989
Edition. Homewood, Ill.: Dow-Jones-Irwin, 1989.
Brickley, J. A., and F. H. Dark. "The Choice of Organizational
Form: The Case of Franchising." Journal of Financial Economics,
July 1987, 401-20.
Carmichael, H. Lorne. "The Agent-Agents Problem: Payments by
Relative Output." Journal of Labor Economics, January 1983, 50-65.
-----. "Efficiency Wage Models of Unemployment - One View."
Economic Inquiry, April 1990, 269-95.
Caves, Richard E., and William F. Murphy. "Franchising: Firms,
Markets, and Intangible Assets." Southern Economic Journal, April
1976, 572-86.
Cheung, Steven N. S. "Transactions Costs, Risk Aversion, and the
Choice of Contractual Form." Journal of Law and Economics, April
1969, 23-42.
Cooper, Russell, and Thomas Ross. "Product Warranties and Double
Moral Hazard." Rand Journal of Economics, Spring 1985, 103-13.
Eswaran, M., and A. Kotwal. "A Theory of Contractual Structure
in Agriculture." American Economic Review, June 1985, 352-67.
Gallini, N., and N. Lutz. "Dual Distribution and Royalty Fees in
Franchising." Journal of Law, Economics, and Organization, October
1992, 471-501.
International Franchise Association Educational Foundation, Inc.
Franchising in the Economy 1988-1990. Washington D.C.: Horwath
International, 1990.
Klein, Benjamin. "Transaction Cost Determinants of
'Unfair' Contractual Arrangements." American Economic
Review, March 1980, 356-62.
Klein, Benjamin, and Keith Leffler. "The Role of Market Forces
in Assuring Contractual Performance." Journal of Political Economy,
August 1981, 615-41.
Klein, Benjamin, and Lester Salt. "The Law and Economics of
Franchise Tying Contracts." The Journal of Law and Economics, May
1985, 345-62.
Lafontaine, Francine. "Agency Theory and Franchising: Some
Empirical Tests." Rand Journal of Economics, Summer 1992, 263-83.
Lal, Rajiv. "Improving Channel Coordination through
Franchising." Marketing Science, Fall 1990, 299-318.
Lang, Kevin, and Shulamit Kahn. "Efficiency Wages Models of
Unemployment: A Second View." Economic Inquiry, April 1990,
296-306.
Lazear, Edward. "Why Is There Mandatory Retirement?"
Journal of Political Economy, December 1979, 1261-84.
Leffler, K., and R. Rucker. "Transactions Costs and the
Efficient Organization of Production: A Study of Timber-Harvesting
Contracts." Journal of Political Economy, October 1991, 1060-87.
Mann, Duncan, and Jennifer Wissink. "Money-back Contracts with
Double Moral Hazard." Rand Journal of Economics, Summer 1988,
285-92.
Martin, Robert. "Franchising and Risk Management." American
Economic Review, December 1988, 954-68.
Mathewson, G. Frank, and Ralph A. Winter. "The Economics of
Franchise Contracts." The Journal of Law and Economics, October
1985, 503-26.
Minkler, A., and T. Park. "Asset Specificity and Vertical
Integration in Franchising." Review of Industrial Organization,
August 1994, 409-23.
Norton, Seth W. "An Empirical Look at Franchising as an
Organizational Form." Journal of Business, April 1988, 197-218.
Rubin, Paul H. "The Theory of the Firm and the Structure of the
Franchise Contract." The Journal of Law and Economics, April 1978,
223-33.
Schmidt, Torsten. "An Analysis of Intrabrand Competition in the
Franchise Industry." Review of Industrial Organization, June 1994,
293-310.
Scott, Frank A. "Franchising vs. Company Ownership as a Decision
Variable of the Firm." Review of Industrial Organization, February
1995, 69-81.
Sen, Kabir. "The Use of Initial Fees and Royalties in
Business-format Franchising." Managerial and Decision Economics,
March 1993, 175-90.
Smith, C., and L. Wakeman. "Determinants of Corporate Leasing
Policy." The Journal of Finance, July 1985, 895-910.
Stiglitz, Joseph. "Incentives and Risk-Sharing in
Share-cropping." Review of Economic Studies, April 1974, 219-55.
U.S. Department of Commerce. Franchising in the Economy 1986-1988.
Washington D.C.: U.S. Department of Commerce, 1988.
Wimmer, Bradley S. "An Analysis of Franchising Compensation
Arrangements." Ph.D. dissertation, University of Kentucky, 1992.
Bradley S. Wimmer: Industry Economist, Federal Communications
Commission, Washington, D.C. Phone 1-202-418-1847, Fax 1-202-418-1567
E-mail bwimmer@fcc.gov
John E. Garen: Professor, University of Kentucky, Lexington Phone
1-606-257-3581, Fax 1-606-323-1920 E-mail jgaren@pop.uky.edu